Electronegativity Calculator
Two fluorines are the most electronegative pair in the table and form a perfectly non-polar bond.
The formula
percent ionic = 100 (1 - exp(-0.25 * (difference)^2))
A difference, not a value
Electronegativity on its own says nothing about a bond. What matters is the difference between the two atoms: two fluorines have the highest electronegativity in the table and form a perfectly non-polar bond, because the difference is zero.
No bond is fully ionic
Pauling's expression approaches 100 % only as the difference grows without limit. The widest difference anywhere in the table is francium against fluorine at 3.28, worth about 93 % — though caesium-fluorine at 3.19 and 92 % is the pair usually quoted, because francium's electronegativity is barely measurable and its longest-lived isotope survives 22 minutes. Sodium chloride, the textbook ionic compound, is about 71 % ionic. The ionic-covalent divide is a spectrum with nothing at either end.
The 1.7 boundary is a convention
The familiar cut-offs — below 0.4 non-polar, above 1.7 ionic — are teaching conventions, chosen because 1.7 happens to fall near 50 % ionic character on Pauling's curve. They are not measurements and nothing physical changes as you cross them.
Models, and where they stop
Slater's rules, the Kapustinskii equation and Pauling's ionic-character expression are all empirical fits, not derivations. They were built to reproduce measured numbers with arithmetic simple enough to do by hand, and they succeed at that within perhaps five or ten per cent. Where a page here quotes one, it says which — a figure from a fitted rule and a figure from a Born-Haber cycle are not the same kind of thing, even when they agree.